Files
DmitryandClaude Sonnet 5 a3fe8031fe feat: migrate all managed LXC provisioning to OpenTofu (blue-green)
Blue-green: a new container is created beside the old one, data is copied, the
IP is moved onto it, and the old container is kept stopped as rollback for at
least a week. Keeping the IP means only the VMID changes, and its consumers
(backup jobs, backup audit) already derive it from the registry.

Batch 1 (2026-09-02): emergency-bot 148->151, docker-test 145->152,
gitea 141->153, vaultwarden 140->154, monitoring 146->155, gyro 150->156,
grimmory 149->157.
Batch 2 (2026-09-03): adguard 144->158, mihomo 143->159, ovpn-mini 132->160.
All migratable LXC are now provisioner: tofu. hermes-ai (frozen) and pbs stay.

- tofu/services.tf + tofu/svc-*.tf: one resource per service, reproducing the
  pct-config etalon. /dev/fuse -> features.fuse; /dev/net/tun ->
  device_passthrough (first live use on mihomo and ovpn-mini); gitea bind mount
  -> datastore volume (data finally reaches PBS); console { type = "shell" }
  declared explicitly (provider tracks cmode there).
- services.yml: vmid + provisioner: tofu for every migrated service; features
  strings and device notes updated to the tofu representation; also drops the
  memoir-bot entry and adds homelab_reverse_proxy_image/_unit.
- pve-*.yml: configuration play target is `{{ pve_config_target | default(...) }}`
  so it can run against <name>-new on a temp address (a bare --limit zeroes the
  play instead of retargeting it). Container-creation plays are gated behind
  `provisioner != 'tofu'` / `pve_provisioning_enabled` (meta: end_play), so a
  stray run cannot pct start a stopped OLD VMID on a live IP. Override for
  intentional legacy rollback: -e pve_<svc>_legacy_provisioning_enabled=true.
- ssh_config: drop memoir-bot; ovpn-mini gets ProxyJump none (a jump via ru-vps
  would route through the very tunnel ovpn-mini terminates).
- gyro.yml / uptime-kuma.yml: same pve_config_target override.
- roles/uptime_kuma: only freeze homelab-monitoring when the unit actually
  exists (a fresh blue-green container never had it).
- offsite-restic-yadisk.yml: the gitea restic profile now runs inside the LXC
  (hosts: gitea), since the bind-mount host path is gone after the volume move;
  lost+found excluded (unreadable in an unprivileged LXC, restic exit 3).

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_012uoq5AVK8mkBgg83Mq6o5V
2026-09-03 07:05:46 +03:00

268 lines
9.3 KiB
YAML

---
- name: Create AdGuard Home LXC on mini-pc
hosts: mini-pc
gather_facts: false
vars:
# После переезда на OpenTofu (tofu/svc-adguard.tf, VMID 158) этот play —
# legacy: он таргетит СТАРЫЙ VMID 144 и его handler сделал бы pct start 144,
# подняв остановленный откат на боевом адресе. Пропускаем, пока реестр
# говорит provisioner: tofu. Для намеренного legacy rollback:
# -e pve_adguard_legacy_provisioning_enabled=true
pve_adguard_legacy_provisioning_enabled: >-
{{ homelab_services['adguard'].provisioner != 'tofu' }}
adguard_vmid: 144
adguard_hostname: adguard
adguard_ip: 192.168.1.28/24
adguard_gateway: 192.168.1.1
adguard_rootfs: local-lvm:8
adguard_ostemplate: local:vztmpl/debian-13-standard_13.1-2_amd64.tar.zst
adguard_pubkey_file: ~/.ssh/id_ed25519_homelab.pub
adguard_image: adguard/adguardhome:v0.107.78@sha256:2c127294fa5f96151d9d3a433fb9d66c17e4d18cf698c2b04372a80e26fdd26f
handlers:
- name: restart adguard lxc
ansible.builtin.shell: "pct stop {{ adguard_vmid }} || true; pct start {{ adguard_vmid }}"
changed_when: true
pre_tasks:
- name: Skip legacy AdGuard provisioning after cutover
ansible.builtin.meta: end_play
when: not (pve_adguard_legacy_provisioning_enabled | bool)
tasks:
- name: Check if AdGuard LXC exists
ansible.builtin.command: "pct config {{ adguard_vmid }}"
register: adguard_pct_config
changed_when: false
failed_when: false
- name: Refuse to modify a foreign VMID {{ adguard_vmid }}
ansible.builtin.assert:
that:
- adguard_pct_config.rc != 0 or adguard_existing_hostname == adguard_hostname
fail_msg: VMID {{ adguard_vmid }} already exists and is not the AdGuard container.
vars:
adguard_existing_hostname: >-
{{ adguard_pct_config.stdout_lines
| select('match', '^hostname: ')
| map('regex_replace', '^hostname: ', '')
| first
| default('') }}
- name: Install AdGuard LXC SSH public key on PVE host
ansible.builtin.copy:
dest: /tmp/adguard-lxc.pub
owner: root
group: root
mode: "0600"
content: "{{ lookup('file', adguard_pubkey_file) }}\n"
when: adguard_pct_config.rc != 0
- name: Create AdGuard LXC
ansible.builtin.command: >-
pct create {{ adguard_vmid }} {{ adguard_ostemplate }}
--hostname {{ adguard_hostname }}
--rootfs {{ adguard_rootfs }}
--cores 1
--memory 512
--swap 512
--net0 name=eth0,bridge=vmbr0,gw={{ adguard_gateway }},ip={{ adguard_ip }},firewall=1
--nameserver 1.1.1.1
--unprivileged 1
--features nesting=1,keyctl=1
--onboot 1
--startup order=40
--cmode shell
--ssh-public-keys /tmp/adguard-lxc.pub
when: adguard_pct_config.rc != 0
- name: Start AdGuard LXC
ansible.builtin.command: "pct start {{ adguard_vmid }}"
register: adguard_pct_start
changed_when: adguard_pct_start.rc == 0
failed_when: adguard_pct_start.rc not in [0, 255]
- name: Allow FUSE device in AdGuard LXC config
ansible.builtin.lineinfile:
path: "/etc/pve/lxc/{{ adguard_vmid }}.conf"
line: "lxc.cgroup2.devices.allow: c 10:229 rwm"
state: present
notify: restart adguard lxc
- name: Bind mount FUSE device in AdGuard LXC config
ansible.builtin.lineinfile:
path: "/etc/pve/lxc/{{ adguard_vmid }}.conf"
line: "lxc.mount.entry: /dev/fuse dev/fuse none bind,create=file"
state: present
notify: restart adguard lxc
- name: Apply pending LXC config changes
ansible.builtin.meta: flush_handlers
- name: Wait for AdGuard SSH through ru-vps
ansible.builtin.wait_for_connection:
timeout: 120
delegate_to: adguard
vars:
ansible_become: false
- name: Configure AdGuard Home inside LXC
# Таргет переопределяем ради blue-green переезда на OpenTofu: во время
# миграции этот play нужно прогнать против нового контейнера на временном
# адресе. Просто `--limit adguard-new` для этого НЕ годится — лимит
# пересекается с паттерном play и даёт ноль хостов, а не перенацеливание
# (проверено 2026-09-02 через --list-hosts на всех плейбуках).
# Использовать вместе с --limit, чтобы play создания контейнера отсеялся:
# ansible-playbook playbooks/pve-adguard.yml \
# -e pve_config_target=adguard-new --limit adguard-new
# По умолчанию поведение не меняется.
hosts: "{{ pve_config_target | default('adguard') }}"
gather_facts: true
vars:
ansible_become: false
adguard_image: adguard/adguardhome:v0.107.78@sha256:2c127294fa5f96151d9d3a433fb9d66c17e4d18cf698c2b04372a80e26fdd26f
tasks:
- name: Install Docker packages
ansible.builtin.apt:
name:
- docker.io
- fuse-overlayfs
- ca-certificates
- curl
- dnsutils
state: present
update_cache: true
- name: Ensure Docker config directory exists
ansible.builtin.file:
path: /etc/docker
state: directory
owner: root
group: root
mode: "0755"
- name: Configure Docker storage driver for unprivileged LXC
ansible.builtin.copy:
dest: /etc/docker/daemon.json
owner: root
group: root
mode: "0644"
content: |
{
"storage-driver": "fuse-overlayfs"
}
register: adguard_docker_daemon_config
- name: Enable Docker service
ansible.builtin.systemd:
name: docker
state: "{{ 'restarted' if adguard_docker_daemon_config.changed else 'started' }}"
enabled: true
- name: Ensure AdGuard data directories exist
ansible.builtin.file:
path: "{{ item }}"
state: directory
owner: root
group: root
mode: "0750"
loop:
- /opt/adguard/work
- /opt/adguard/conf
- name: Check if the configured AdGuard image is present
ansible.builtin.command: "docker image inspect {{ adguard_image }}"
register: adguard_image_inspect
changed_when: false
failed_when: false
- name: Pull the configured AdGuard image
ansible.builtin.command: "docker pull {{ adguard_image }}"
when: adguard_image_inspect.rc != 0
register: adguard_image_pull
changed_when: true
- name: Install AdGuard Home systemd unit
ansible.builtin.copy:
dest: /etc/systemd/system/adguard.service
owner: root
group: root
mode: "0644"
content: |
[Unit]
Description=AdGuard Home container
After=docker.service
Requires=docker.service
[Service]
Restart=always
RestartSec=10
ExecStartPre=-/usr/bin/docker rm -f adguard
ExecStart=/usr/bin/docker run --rm --name adguard --pull never \
-p 53:53/tcp -p 53:53/udp -p 80:80/tcp -p 3000:3000/tcp \
-v /opt/adguard/work:/opt/adguardhome/work \
-v /opt/adguard/conf:/opt/adguardhome/conf \
{{ adguard_image }}
ExecStop=/usr/bin/docker stop adguard
[Install]
WantedBy=multi-user.target
register: adguard_unit
- name: Remove obsolete AdGuard post-start validation unit
ansible.builtin.file:
path: /etc/systemd/system/adguard-post-start.service
state: absent
register: adguard_post_start_unit_removed
- name: Reload systemd when AdGuard units change
ansible.builtin.systemd:
daemon_reload: true
when: adguard_unit.changed or adguard_post_start_unit_removed.changed
- name: Enable and start AdGuard Home
ansible.builtin.systemd:
name: adguard
state: "{{ 'restarted' if adguard_unit.changed or adguard_image_pull.changed else 'started' }}"
enabled: true
- name: Wait for AdGuard HTTP root
ansible.builtin.uri:
url: http://127.0.0.1/
status_code: [200, 302]
return_content: false
register: adguard_http_root
retries: 24
delay: 5
until: adguard_http_root.status in [200, 302]
- name: Verify AdGuard DNS over UDP
ansible.builtin.command:
argv:
- dig
- "@127.0.0.1"
- localhost
- A
- +time=2
- +tries=1
- +short
register: adguard_dns_udp
changed_when: false
retries: 12
delay: 5
until: adguard_dns_udp.rc == 0 and '127.0.0.1' in adguard_dns_udp.stdout
- name: Verify AdGuard DNS over TCP
ansible.builtin.command:
argv:
- dig
- "@127.0.0.1"
- localhost
- A
- +tcp
- +time=2
- +tries=1
- +short
register: adguard_dns_tcp
changed_when: false
retries: 12
delay: 5
until: adguard_dns_tcp.rc == 0 and '127.0.0.1' in adguard_dns_tcp.stdout